**** BEGIN LOGGING AT Fri Jul 05 02:59:56 2019 Jul 05 07:24:23 Morning! Jul 05 07:24:35 Tofe: Almost done re-incorporating all those sleepd patches Jul 05 07:24:48 Then need to test to see if it builds & how it behaves runtime Jul 05 07:38:37 Morning! Jul 05 07:38:44 Herrie: if you want I can have a look Jul 05 07:38:53 Tofe: Well I have the patches mostly there Jul 05 07:38:57 Let me do a build check first Jul 05 07:39:00 See if there's havoc Jul 05 07:39:42 Not really pleased with how it was hacked together previously to be honest Jul 05 07:40:00 It replaced the synchonous by asynchronous from what I can tell. Jul 05 07:40:02 :) it's a good opportunity to improve it a bit Jul 05 07:40:12 Which is fine for our targets, but ideally you'd support both Jul 05 07:40:22 But there's quite some of the display state etc there Jul 05 07:40:30 So that explains the backlight behavior I'd say Jul 05 08:03:23 Tofe: This is what I ported from webOS-ports/master to webOS-ports/webOS-OSE: https://github.com/webOS-ports/sleepd/pull/4 Jul 05 08:03:30 As you can see some bits aren't very pretty Jul 05 08:03:45 There's a number of times: #if 0 Jul 05 08:04:06 In general it builds now, but I haven't runtime tested it yet Jul 05 08:41:45 Let me do build locally for tenderloin Jul 05 08:51:16 for rosy I think I'll have to reflash from scratch Jul 05 08:51:50 there's plenty of space, and I don't see why recovery would crash when doing a simple extraction Jul 05 08:53:04 Poking the LG guys a bit for PulseAudio upgrade Jul 05 08:53:23 They did some custom resampler at their end which seems to date back from legacy Jul 05 08:53:34 But they're stuck on 9.0 now, while Thud & Warrior are at 12.2 Jul 05 08:53:44 The patches shouldn't be rocket science to add, but still ;) Jul 05 08:57:49 Well, like for so many pieces, they forked it and now it adds efforts on their end to upgrade... While our pulseaudio works fine Jul 05 08:58:45 Though we don't have an evolved media stack like them, so a resampler was probably a need Jul 05 09:07:10 Most additions are pretty clean though Jul 05 09:15:45 https://github.com/Herrie82/pulseaudio-webos/compare/master...webOS-OSE/modifications Jul 05 09:15:56 I see some room for improvements here and there, but not a whole lot Jul 05 09:16:07 These patches could actually fairly easily be upstreamed as well Jul 05 09:36:29 most of is can be in a separate repo: the palm policy module, the conf files... only the resampler itself would need to be upstreamed Jul 05 09:37:45 their #ifdef PALM_UNMODIFIED_CODE is worrying though, like their new code may not be working well in the general scenario Jul 05 09:46:17 Yes that's what I also thought Jul 05 09:47:16 Anyway why would you use #ifdef PALM_UNMODFIED_CODE ? They could just as well use the HAVE_PALM_RESAMPLER somehow Jul 05 10:35:25 I think they want to check regularly if a bug is caused by their new code or not Jul 05 11:17:55 Tofe: Found a little bug in webos-telephonyd migration (typo in filename) causing some havoc Jul 05 11:18:10 This might be the cause of some of the FirstUse issues Jul 05 11:18:15 Testing a fix now Jul 05 11:19:38 Seems I named the role file for sysbus incorrectly Jul 05 11:19:44 So it didn't get put in the right place etc Jul 05 11:20:42 Leading to LS2 permission errors :S Jul 05 11:25:05 OK now getting some different errors, but I know how to solve those;) Jul 05 12:14:22 Tofe: This will solve some of the error messages. I haven't tested it on device yet to see if it actually fixes stuff, but it won't hurt to fix it: https://github.com/webOS-ports/webos-telephonyd/pull/15 Jul 05 12:14:42 There are a few more calls with LS_REQUIRES_SECURITY errors in the logs Jul 05 12:15:03 I'll "fix" those too. Just by allowing their calls by "public" and "private" for now Jul 05 12:15:21 These happen with components that are already on ACG, but the components that are calling them not yet for example Jul 05 12:21:07 Herrie: thanks, merged Jul 05 12:21:22 There will be a few more by the looks of it Jul 05 12:21:25 Just going through logs Jul 05 12:21:33 This might solve some of the WiFi/FirstUse behavior Jul 05 12:22:23 It won't hurt to fix them anyway ;) Jul 05 12:22:59 Even if it doesn't fix anything in the end Jul 05 12:27:35 OK this kills another 2 entries in the logs: https://github.com/webOS-ports/webos-telephonyd/pull/16 Jul 05 15:21:17 Tofe: The fixes don't see to solve a whole lot of issues, but well ;0 Jul 05 15:21:18 ;) Jul 05 15:21:30 I think the sleepd actually does improve stuff ;) Jul 05 15:22:22 My backlight etc behaves better Jul 05 15:25:05 At least on Tenderloin Jul 05 15:25:13 Need to test a bit on Hammerhead as well I guess Jul 05 16:44:57 ok, so... let's merge Jul 05 16:47:07 Yup ;) Just some cleanup mainly, but also needs to happen ;) Jul 05 16:47:39 Was a bit surprised I needed to have the com,palm,app.* in the Prefs tab as well, but well, this works around it for now ;) Jul 05 18:08:52 I would say this should be stable enough for a release in general Jul 05 18:09:09 Let me bump the SRCREV's and kick off some new builds in a minute Jul 05 18:13:50 Just not sure how to force only tenderloin on the specific libhybris SRCREV Jul 05 18:21:19 OK Jenkins jobs are running Jul 05 18:41:11 in a libhybris bbappend near meta-hp, just put SRCREV_tenderloin = "..." Jul 05 18:41:39 Ican do that Jul 05 18:47:55 Herrie: https://github.com/shr-distribution/meta-smartphone/pull/108 Jul 05 18:53:42 Well that's what I tried locally but had it in our meta-webos-ports bbappend Jul 05 18:54:08 Had some issues locally not sure what caused it **** ENDING LOGGING AT Sat Jul 06 02:59:56 2019